MAJOR ARTICLE

Aspirin Use for Primary and Secondary

Prevention in Human Immunodeciency Virus
(HIV)-Infected and HIV-Uninfected Patients
Sujit Suchindran,1,2 Susan Regan,3 James B. Meigs,3 Steven K. Grinspoon,4 and Virginia A. Triant2,3
1

Division of Infectious Diseases, Brigham and Womens Hospital, Boston, Massachusetts; 2Division of Infectious Diseases, 3Division of General Internal
Medicine, and 4Program in Nutritional Metabolism, Massachusetts General Hospital, Boston

Downloaded from http://ofid.oxfordjournals.org/ by guest on October 20, 2014

Background. Human immunodeciency virus (HIV) infection is associated with increased risk of myocardial
infarction (MI). The use of aspirin for primary and secondary MI prevention in HIV infection has not been extensively studied.
Methods. We performed a cross-sectional study of 4037 patients infected with HIV and 36 338 demographicsmatched control patients in the Partners HealthCare System HIV cohort. We developed an algorithm to ascertain
rates of nonepisodic acetylsalicylic acid (ASA) use using medication and electronic health record free text data. We
assessed rates of ASA use among HIV-infected and HIV-uninfected (negative) patients with and without coronary
heart disease (CHD).
Results. Rates of ASA use were lower among HIV-infected compared with HIV-uninfected patients (12.4% vs
15.3%, P < .001), with a relatively greater difference among patients with 2 CHD risk factors (22.1% vs 42.4%,
P < .001). This nding was present among men and among patients in the 3039 and 4049 age groups. Among
patients with prevalent CHD using ASA for secondary prevention, rates of ASA use were also lower among HIVinfected patients compared with HIV-uninfected patients (51.6% vs 65.4%, P < .001).
Conclusions. Rates of ASA use were lower among HIV-infected patients compared with controls, with a greater
relative difference among those with elevated CHD risk and those with known CHD. Further studies are needed to
investigate the optimal strategies for ASA use among patients infected with HIV.
Keywords. aspirin; coronary disease; HIV; prevention; primary care.

Because treatment of human immunodeciency virus

(HIV) has become more widespread and effective,
there has been an increase in overall life expectancy
[1, 2] but also a relative increase in noninfectious complications [3, 4]. Human immunodeciency virus infection has been associated with increased risk of coronary

Received 2 May 2014; accepted 13 June 2014.

Correspondence: Virginia A. Triant, MD, MPH, Divisions of Infectious Diseases
and General Internal Medicine, Massachusetts General Hospital, 50 Staniford St.,
9th Floor, Boston, MA 02114 (vtriant@mgh.harvard.edu).
Open Forum Infectious Diseases
The Author 2014. Published by Oxford University Press on behalf of the Infectious
Diseases Societyof America. This is an Open Access article distributed under the terms
of the Creative Commons Attribution-NonCommercial-NoDerivs licence (http://
creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial
reproduction and distribution of the work, in any medium, provided the original work
is not altered or transformed in any way, and that the work is properly cited. For
commercial re-use, please contact journals.permissions@oup.com.
DOI: 10.1093/od/ofu076

heart disease (CHD) and myocardial infarction (MI) in

multiple studies across different care settings [59].
Investigation has been increasingly directed towards
understanding the mechanism of CHD in HIV-infected
patients, with suggested links including higher rates of
traditional MI risk factors [10], use of antiretroviral
(ARV) medication [11], chronic inammation [12],
and immune dysfunction [13]. Despite the increased
understanding of CHD pathophysiology in HIV, little
is known about CHD prevention in this group. In particular, it is unclear whether traditional methods of
CHD risk reduction are used in a similar manner or
provide a similar benet in HIV-infected (negative) patients compared with HIV-uninfected patients. Aspirin
(acetylsalicylic acid [ASA]) is an inhibitor of platelet aggregation that has been shown to be effective in the primary [1416] and secondary [17] prevention of MI,
although few studies have investigated its use in patients
Aspirin Use and HIV

OFID

with HIV. Our primary objective was to compare rates of ASA

use for both primary and secondary prevention between patients infected with HIV and matched control patients.
METHODS
Study Design

Acetylsalicylic Acid Rate Measurement

Acetylsalicylic acid use rates were ascertained based on an algorithm developed for this study using a combination of medication data and EHR free text notes. Medication data included
inpatient and outpatient prescriptions. All EHR clinical notes
were loaded into an SQL-Server database, which searched for
evidence of ASA use using a strategy that located the terms aspirin, ASA, acetylsalicylic acid, and Aggrenox and that
eliminated cases in which the surrounding text indicated nonuse (eg, mentions of allergies or instructions to avoid ASA).
Other brand names of ASA and possible misspellings of ASA
were initially included but yielded no search results. We established a criterion for nonepisodic ASA use combining the medication and note data. Patients were classied as ASA users if
they had at least: (1) 2 prescriptions, (2) notes with ASA text
on 2 or more days, or (3) a prescription and a note. Patients
were classied as using ASA for primary prevention if the
2

OFID

Suchindran et al.

Statistical Analysis

For patients without known CHD at start of observation, the

frequencies of MI risk factors and HIV parameters (when applicable) were calculated in the HIV-infected and HIV-uninfected
groups, stratied by ASA use. The 2 test was used to compare
categorical baseline characteristics between ASA users and nonusers in the both the HIV-infected and HIV-uninfected groups.
We measured rates of nonepisodic ASA use using the described
algorithm in HIV-infected and HIV-uninfected patients stratied by age group, CHD risk (high risk dened as 2 risk factors
vs low risk dened as 01 risk factors, with risk factors including hypertension, dyslipidemia, diabetes, and smoking), and
CD4 nadir in HIV-infected patients. We repeated the analysis
for HIV-infected and HIV-uninfected patients with CHD before the start of observation, with the exception of stratication
by CD4 nadir.
RESULTS
Demographic Characteristics

Demographic and clinical characteristics of patients without

known CHD are outlined in Table 1. Among both HIV-infected

Downloaded from http://ofid.oxfordjournals.org/ by guest on October 20, 2014

To assess rates of ASA use for primary prevention, we conducted a cross-sectional study of 4037 patients infected with HIV
and 36 338 HIV-uninfected patients without known CHD
from the Research Patient Data Registry, a clinical care database
of patients from Massachusetts General Hospital and Brigham
and Womens Hospital, comprising the Partners HealthCare
System in Boston, MA. Data from inpatient and outpatient visits are included in the database, as previously described [18].
Human immunodeciency virus ascertainment was previously
validated [19] and was established based on International Classication of Diseases, 9th Revision, Clinical Modication (ICD9-CM) codes 042 or V08 or corresponding electronic health
record (EHR) codes, with initial code occurring before MI if
one did occur. The HIV-uninfected group was matched in a
10:1 ratio on age, gender, and race. The time period of analysis
started no later than January 1, 2000, the rst ICD-9-CM code for
HIV when applicable, the rst encounter, or the patients 18th
birthday. The analysis period ended no earlier than December
31, 2009, rst MI, or last encounter.
We also assessed rates of ASA use among HIV-infected and
HIV-uninfected patients with known CHD. We evaluated 410
HIV-infected and 3366 HIV-uninfected patients with prevalent
CHD, diagnosed based on ICD-9-CM codes 410414 before
January 1, 2008 and before HIV code, if applicable. For secondary prevention, the analysis period started at the CHD diagnosis
date and ended at last encounter or December 31, 2009, whichever was earlier.

notes or prescriptions were dated at least 30 days before the

MI event. We classied patients as ASA nonusers if they had no
notes and no prescriptions. Those with only one note or one
prescription were excluded (n = 3329 of 40 375). We validated
this algorithm using physician chart review of 207 charts, selected at random equally from HIV-infected and HIV-uninfected
patients. The algorithm yielded a sensitivity of 73%, specicity
of 83%, and area under the receiver-operating characteristics
curve (AUC) of 0.78. When stratied by HIV status, AUC was
0.85 for HIV-infected and 0.76 for HIV-uninfected patients. For
the secondary prevention analysis, patients were required to
meet criteria for ASA use after the initial CHD diagnosis but
before the end of the analysis period.
Patient characteristics including hypertension, diabetes mellitus, dyslipidemia, thromboembolic disease (deep venous thrombosis or pulmonary embolism), atrial brillation, and hepatitis C
virus were dened using ICD-9-CM coding (see Table 1 for relevant codes). Additional exposure ascertainment included ICD9-CM code 571 for chronic liver disease, 578 for gastrointestinal
bleed, and 531534 for ulcer disease. We obtained smoking status
using a validated algorithm developed based on a natural language processing tool [20, 21]. For HIV-infected patients, patient
characteristics measured included CD4 count, HIV RNA, and
history of ARV use, subclassied as protease inhibitor (PI) use,
nonnucleoside reverse-transcriptase inhibitor use, or nucleoside
reverse-transcriptase inhibitor use. Recent CD4 and HIV RNA
were dened as the latest laboratory measurement before the
end of the analysis period. All patient characteristics were restricted to occurring before MI, if one occurred.

Table 1.

Baseline Characteristics of Study Participants Without Known CHDa

HIV-Infected Patients
ASA Use
n = 458

No ASA Use
n = 3240

HIV Uninfected Patients

P Value

ASA Use
n = 5089

No ASA Use
n = 28259

P Value

<.001
.051

47.9 (11.6)
1327 (26.1)

40.1 (11.4)
10321 (36.5)

<.001
<.001

1071 (21.1)
2983 (58.6)

6312 (22.3)
13936 (49.3)

751 (14.8)

5312 (18.8)

95 (1.9)
189 (3.7)

845 (3.0)
1854 (6.6)

Demographics
Age, mean (SD)
Women

48.0 (11.6)
123 (26.9)

40.2 (10.1)
1016 (31.4)

104 (22.7)
249 (54.4)

680 (21.0)
1632 (50.4)

Hispanic

82 (17.9)

565 (17.4)

Other
Unknown

10 (2.2)
13 (2.8)

99 (3.1)
264 (8.2)

Race
African-American
Caucasian

Duration of observation, mean years (SD)

.001

<.001

5.1 (3.4)

<.001

6.8 (3.3)

4.1 (3.6)

<.001

290 (63.3)

813 (25.1)

<.001

3219 (63.3)

4650 (16.5)

<.001

Diabetes mellitus

185 (40.4)

486 (15.0)

<.001

1547 (30.4)

1606 (5.7)

<.001

Dyslipidemia
Smoking

294 (64.2)
306 (66.8)

1128 (34.8)
1695 (52.3)

<.001
<.001

3168 (62.3)
2584 (50.8)

4681 (16.6)
7657 (27.1)

<.001
<.001

Atrial fibrillation

43 (9.4)

44 (1.4)

<.001

462 (9.1)

351 (1.2)

<.001

34 (7.4)
146 (31.9)

61 (1.9)
656 (20.3)

<.001
<.001

172 (3.4)
186 (3.7)

195 (0.7)
430 (1.5)

<.001
<.001

Recent CD4 count (cells/mm3), median (IQR)

Nadir CD4 count (cells/mm3), median (IQR)

390 (218621)
186 (53340)

431 (253648)
221 (89389)

.096
.006

Nadir CD4 count <200

152 (53.7)

893 (45.8)

.013

HIV RNA recent, median log copies/mL (IQR)c

HIV RNA <400 (undetectable)

4.0 (2.94.8)
186 (67.9)

4.0 (2.74.8)
1141 (61.9)

.699
.055

ARV medication used

278 (60.7)

1501 (46.3)

<.001

PI use
NNRTI use

194 (42.4)
172 (37.6)

981 (30.3)
799 (24.7)

<.001
<.001

NRTI use

265 (57.9)

1432 (44.2)

<.001

Thromboembolic disease
HCV infection
HIV Parameters

Abbreviations: ARV, antiretroviral; ASA, acetylsalicylic acid; CHD, coronary heart disease; HCV, hepatitis C virus; HIV, human immunodeficiency virus; IQR,
interquartile range; NNRTI, nonnucleoside reverse transcriptase inhibitor; PI, protease inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; SD, standard
deviation.
a

Unless indicated otherwise, data are expressed in number (percentage) of patients.

ICD-9-CM codes used to identify diagnoses: hypertension, 401; diabetes mellitus, 250; dyslipidemia, 272; atrial fibrillation, 427.3; thromboembolic disease, 415.1
(pulmonary embolism) or 453.4 (deep vein thrombosis); HCV infection, 070.5, 070.7, 070.41, or 070.44; CHD, 410414.

Among those with detectable viral load.

ARV use is categorized as ever/never.

and HIV-uninfected patients, ASA users were older and had

signicantly higher rates of traditional MI risk factors, including
hypertension, diabetes, dyslipidemia, and smoking, although
the relative differences in risk factors between ASA users and
nonusers were greater in HIV-uninfected patients. Human immunodeciency virus-infected ASA users had a signicantly
lower nadir CD4 count and a higher proportion of patients
on ARV medications. Patients in the HIV-infected group also
had higher rates of chronic liver disease (13.1% vs 3.2%), gastrointestinal bleed (11.0% vs 5.5%), and ulcer disease (3.7% vs
1.5%).

Rates of Acetylsalicylic Acid Use

Rates of ASA use are outlined in Table 2. Acetylsalicylic acid use

was lower in HIV-infected compared with HIV-uninfected patients in the overall group (12.4% vs 15.3%, P < .001) and
among men (13.1% vs 17.3%, P < .001), as shown in Figure 1.
In age-stratied analyses, ASA use was signicantly lower
among HIV-infected patients age 3039 and age 4049,
which was the largest of the age subgroups, but it was signicantly higher among HIV-infected men over the age of 70.
Rates of ASA used were signicantly lower for HIV-infected
compared with HIV-uninfected patients in both the high
Aspirin Use and HIV

OFID

Downloaded from http://ofid.oxfordjournals.org/ by guest on October 20, 2014

5.9 (3.1)

Vascular Risk Factorsb

Hypertension

Table 2.

Rates of ASA Use Among HIV and Control Patients Without Known CHDa
Overall

Women

MEN

HIVInfected
n = 3698

HIVUninfected
n = 33 348

P
Value

HIVInfected
n = 1139

HIVUninfected
n = 11 648

P
Value

HIVInfected
n = 2559

HIVUninfected
n = 21 700

P
Value

458 (12.4)

5089 (15.3)

<.001

123 (10.8)

1327 (11.4)

.547

335 (13.1)

3762 (17.3)

<.001

1829
3039

16 (3.0)
88 (7.2)

236 (4.2)
1009 (10.0)

.194
.002

7 (3.0)
28 (7.1)

119 (3.9)
265 (7.7)

.491
.657

9 (3.0)
60 (7.3)

117 (4.5)
744 (11.2)

.231
.001

4049

171 (13.6)

1880 (17.8)

<.001

48 (14.2)

446 (14.2)

.983

123 (13.4)

1434 (19.2)

<.001

5059
6069

116 (22.0)
47 (39.5)

1243 (25.1)
496 (34.5)

.116
.271

20 (17.1)
14 (45.2)

271 (20.7)
133 (32.4)

.351
0.147

96 (23.4)
33 (37.5)

972 (26.6)
363 (35.3)

.152
.680

20 (45.5)

225 (36.8)

.250

6 (30.0)

93 (35.8)

.603

14 (58.3)

132 (37.5)

.043

Overall prevalence
Age group

>70
CHD risk factorsb
01
2

200 cells/mm

1687 (6.7)

339 (22.1)

3402 (42.4)

152 (14.6)

131 (11.0)

.037
<.001
.013c

36 (5.2)

505 (5.3)

87 (19.7)

822 (38.5)

44 (14.8)

25 (7.8)

.864
<.001
.006c

83 (5.7)

1182 (7.5)

252 (23.0)

2580 (43.8)

108 (14.5)

106 (12.3)

.011
<.001
.194c

Abbreviations: ASA,acetylsalicylic acid; CHD,coronary heart disease; HIV, human immunodeficiency virus.
a

Unless indicated otherwise, data are expressed in number (percentage) of patients.

CHD risk factors: hypertension, dyslipidemia, diabetes mellitus, smoking.

P value comparing rates of ASA use among HIV-infected by CD4 nadir.

cardiovascular risk (22.1% vs 42.4%, P < .001) and low cardiovascular risk (5.5% vs 6.7%, P = .037) groups, but the relative
difference was greater in patients in the high-risk group, as
shown in Figure 1. The relatively larger difference in the
high-risk group was seen among both women and men.
Among HIV-infected patients, the rates of ASA use were

signicantly higher among those with a nadir CD4 count of

<200 compared with those with a nadir CD4 count of 200,
among the overall group and women. To address temporal
changes in rates of ASA use between HIV-infected and HIVuninfected patients, we compared the distribution of ASA use
start years by HIV status and found the mean start year of

Figure 1. Rates of acetylsalicylic acid (ASA) use in human immunodeciency virus (HIV)-infected vs HIV-uninfected patients in overall group (A), low
coronary heart disease (CHD) risk patients (B), and high CHD risk patients (C). Rates of ASA use are shown in the overall population and within each gender
for HIV-infected and HIV-uninfected patients without known CHD. Rates of ASA use are shown in low and high CHD risk (01 risk factors and 2 risk
factors, respectively) HIV-infected and HIV-uninfected patients without known CHD. Coronary heart disease risk factors assessed were hypertension, dyslipidemia, diabetes, and smoking. P values are for comparison of rates between HIV-infected and HIV-uninfected patients within each group.
4

OFID

Suchindran et al.

Downloaded from http://ofid.oxfordjournals.org/ by guest on October 20, 2014

CD4 count nadir

<200 cells/mm3

119 (5.5)

Table 3.

Rates of ASA Use Among HIV And Control Patients With Known CHDa
Overall

Women

Men

HIVInfected
n = 403

HIVUninfected
n = 3331

P
Value

HIVInfected
n = 110

HIVUninfected
n = 700

P
Value

HIVInfected
n = 293

HIVUninfected
n = 2631

P
Value

208 (51.6)

2179 (65.4)

<.001

49 (44.6)

415 (59.3)

.004

159 (54.3)

1764 (67.1)

<.001

1839
4049

21 (32.3)
82 (50.6)

215 (46.1)
675 (61.9)

.036
.006

9 (33.3)
19 (40.4)

46 (38.3)
127 (57.2)

.628
.036

12 (31.6)
63 (54.8)

169 (48.8)
548 (63.1)

.043
.083

5059

68 (57.6)

754 (72.7)

.001

12 (54.6)

121 (66.9)

.252

56 (58.3)

633 (74.0)

.001

37 (63.8)

535 (72.5)

.156

9 (64.3)

121 (68.4)

.753

28 (63.6)

414 (73.8)

.144

Overall prevalence
Age group

>59
CHD risk factorsb
01

23 (24.7)

309 (36.6)

.023

5 (20.0)

61 (30.8)

.265

18 (26.5)

248 (38.3)

.054

185 (59.7)

1870 (75.2)

<.001

44 (51.8)

354 (70.5)

.001

141 (62.7)

1516 (76.4)

<.001

Abbreviations: ASA, acetylsalicylic acid; CHD, coronary heart disease; HIV, human immunodeficiency virus.
a

Unless indicated otherwise, data are expressed in number (percentage) of patients.

CHD risk factors: hypertension, dyslipidemia, diabetes mellitus, smoking.

DISCUSSION
In this large clinical care cohort comprised of more than 40 000
patients, we found that among patients infected with HIV, ASA
was underused in primary and secondary prevention of MI
compared with HIV-uninfected patients. Using a validated
algorithm to ascertain nonepisodic ASA use, we showed that
rates of ASA use were signicantly lower in patients infected

with HIV compared with matched controls, and that this difference was relatively greater among those at higher cardiovascular risk and those with known CHD.
Prior studies have shown ASA to be under utilized for the
primary prevention of CHD in patients infected with HIV,
with only 17% of patients who met criteria based on US Preventive Services Task Force recommendations receiving ASA [22].
Another study documented less than 2% of HIV-infected
patients using ASA, although it was indicated by guidelines
for approximately 31% [23]. However, neither of these studies
included an HIV-uninfected population. In addition, the
study was not able to compare ASA use in stratied models
with patients comprehensively phenotyped for CVD risk and
separated by gender. Moreover, these studies have not compared ASA use in primary and secondary prevention. We determined rates of ASA use in HIV-infected and HIV-uninfected
patients, with stratication by age, gender, and CHD risk
group. We ensured that there was at least a 30-day gap between
ASA use documentation and initial MI to eliminate cases in
which ASA use was initiated as a consequence of MI. Aspirin
use for primary prevention was lower among HIV-infected patients compared with HIV-uninfected patients overall and
among males, but not among females. There were also striking
differences in rates of ASA use by CHD risk group, with ASA
rates increased 20% in HIV-uninfected vs HIV-infected patients
at low risk but nearly double for patients at high risk. Aspirin use
for secondary prevention of MI was also signicantly lower
among HIV-infected patients compared with HIV-uninfected
patients overall and among both women and men.
The reason for lower rates of ASA use among HIV-infected
patients is unclear. One possible reason for the discrepancy is
the lack of CHD prevention guidelines specic to the HIV
Aspirin Use and HIV

OFID

Downloaded from http://ofid.oxfordjournals.org/ by guest on October 20, 2014

ASA use to be similar (2005 for each group). To address whether contraindications contributed to lower ASA rates in HIV,
we assessed the effect of HIV status on ASA use (using multivariable analysis), controlling for chronic liver disease, ulcer disease, gastrointestinal bleed, and hepatitis C virus. Human
immunodeciency virus remained signicantly associated
with decreased ASA (odds ratio for ASA, .56; 95% condence
interval, .50.63).
Rates of ASA use among patients with prevalent CHD at the
start of observation are outlined in Table 3. Acetylsalicylic acid
use was lower in HIV-infected patients compared with HIVuninfected patients overall (51.6% vs 65.4%, P < .001) and within
each gender (44.6% vs 59.3%, P = .004 among women and 54.3%
vs 67.1%, P < .001 among men). In age-stratied analyses, rates of
ASA use were signicantly lower in the HIV group among patients aged 1839 (32.3% vs 46.1%, P = .036), 4049 (50.6% vs
61.9%, P = .006), and 5059 (57.6% vs 72.7%, P = .001) years
but not among patients older than age 59. Similar lower rates
of ASA use in age stratied analyses were seen in the subgroup
of HIV-infected males. Among all patients with known CHD,
lower rates of ASA use were seen in patients with 01 CHD
risk factors (24.7% vs 36.6%, P = .023) and among patients
with 2 risk factors (59.7% vs 75.2%, P < .001).

OFID

Suchindran et al.

for secondary prevention. The low rates observed among both

HIV-infected and HIV-uninfected patients may reect underuse of this treatment modality, which has been demonstrated
for the general population [28]. However, the underlying reasons for low rates of ASA use for secondary prevention in the
general population remain unknown. There is a clear need for
targeted interventions to increase ASA use in this high-risk
population of both HIV-infected and HIV-uninfected patients.
Indications for and optimal rates of ASA use in patients infected with HIV remain unknown. In the general population,
multiple meta-analyses of randomized trials of primary prevention have shown that ASA was associated with reduction in vascular events, including MI, but not necessarily a reduction in
mortality [15, 16, 29]. The evidence supporting ASA use for secondary prevention of MI in the general population is stronger,
with a meta-analysis showing antiplatelet therapy to be associated with a 22% reduction in the risk of serious vascular event
and reduction in vascular mortality [17]. Whether this data on
the efcacy of ASA use for event reduction applies to patients
infected with HIV remains incompletely understood and is a
focal point for further investigation.
Mechanistic evidence on platelet activation and endovascular
damage in patients infected with HIV-infected suggests that
ASA might play at least a partial role in CHD prevention for
this group. Platelet activation and aggregation are necessary
steps for thrombus formation and arterial occlusion in acute
coronary syndrome [30], and ASA reduces platelet production
of thromboxane A2, thereby decreasing platelet aggregation.
Early evidence suggested that compared with HIV-uninfected
controls, patients infected with HIV had increased platelet activation, as measured by ow cytometric analysis of markers
P-selectin and CD63 [31]. Further study has shown differential
effects between HIV-infected and HIV-uninfected patients in
response to various agonists of platelet aggregation, with unclear effects on the overall association between HIV infection
and platelet aggregation [32]. There are only limited data on
the effects of ASA on platelet aggregation in patients infected
with HIV; however, in vitro data suggest that ASA reduces
platelet aggregation in patients infected with HIV, although in
slightly different ways than for HIV-uninfected patients, and
also reduces a generalized marker of immune activation in
HIV-infected patients [33]. Overall, it remains unclear whether
ASA will have the same effects in HIV-infected patients as in
HIV-uninfected patients given the novel CHD-related risk factors in HIV, which may not be mitigated by ASA. This underlying issue of whether traditional CHD prevention measures
such as ASA or statin therapy have a different role in HIV
infection remains an open question. Furthermore, the role of
novel therapies targeting chronic inammation and immune
activation to decrease CHD risk in HIV infection requires investigation in addition to study of more conventional CHD prevention strategies in HIV populations.

Downloaded from http://ofid.oxfordjournals.org/ by guest on October 20, 2014

population. Although recent HIV primary care guidelines do

provide information on statin use, particularly as they pertain
to drug interactions as well as recommendations for diabetes
screening, they do not provide guidance on starting pharmacologic prevention measures such as ASA or statins [24]. This
omission likely reects a lack of studies of the role of ASA or
statins, specically in HIV. Moreover, CHD prevention strategies for the general population continue to evolve, as evidenced
by recently released cholesterol treatment guidelines [25]. New
CHD risk calculators that underlie the guideline changes merit
validation in the HIV population, given novel risk factors in the
pathogenesis of CHD in HIV infection before formalized adoption of the new cholesterol guidelines in the HIV-infected
population.
This discrepancy in rates of ASA use could also reect a generalized under emphasis by HIV providers on noninfectious
chronic disease complications of HIV in the earlier years included in the time period of the study. As the paradigm of
HIV care in developed countries shifts from preventing acquired immune deciency syndrome-related complications
and treating opportunistic infections to the chronic disease
management of relatively healthier patients, clinical practice
may take time to incorporate the vast number of changes needed. As noted, some of the chronic disease complications are reected in recent HIV primary care guidelines that provide
guidance on topics including diabetes, colon cancer, and bone
density screening [24]. One recent study showed that breast cancer screening was lower among HIV-infected patients compared
with HIV-uninfected patients and that HIV-infected patients
were less likely to receive a colonoscopy purely for colorectal
cancer screening purposes [26]. Another study showed low
rates of successful treatment of hypertension among men and
women infected with HIV [27]. Human immunodeciency
virus providers must continue to be informed of updates in primary care in addition to their specialty knowledge, including
advances in HIV care and an expanded array of ARV medications. An additional reason for lower rates of ASA use in patients infected with HIV could be relatively higher rates of
conditions that might increase bleeding risk. In our cohort, patients infected with HIV had higher rates of chronic liver disease, gastrointestinal bleed, and ulcer disease, any of which
could represent a relative contraindication to ASA use. The
nding that HIV status remained associated with decreased
rates of ASA use when controlling for conditions that predispose to bleeding suggests that these factors alone are not likely
to explain the lower rates of ASA use observed in HIV.
The gap in ASA use for secondary prevention in HIV-infected
patients compared with HIV-uninfected patients may also be
accounted for by a lack of specic ASA-related guidelines or
an under emphasis by HIV providers. However, the overall
low rates in both HIV-infected and HIV-uninfected groups
are striking considering the strong evidence favoring ASA use

Strengths of our study include the use of a well validated,

large data registry, previously shown to perform well in the
characterization of MI, CHD risk rates, and stroke in HIVinfected vs HIV-uninfected populations [6, 13, 34]. Using this
database, we were able to generate a tailored control group
matched to the HIV cohort, from the same background population. Therefore, for the rst time, we were able to assess rates
of ASA use in patients infected with HIV in comparison with
matched control patients. We also developed and validated an
algorithm utilizing EHR free text search to accurately identify
ASA use, an exposure that can be difcult to ascertain using
prescription medication data alone. Limitations of our study
include the incomplete ability to assess dosing of and adherence
to ASA. We were also unable to evaluate over-the-counter ASA
use if it was not explicitly mentioned in provider notes. Nonetheless, we expect similar limitations for both HIV-infected
patients and matched control patients.

In summary, our study demonstrates lower rates of ASA use

among patients infected with HIV compared with controls, particularly in patients at high underlying cardiovascular risk and
those with known CHD. Our ndings underscore the need for
the development of evidence-based, tailored CHD prevention
strategies in the HIV population.
Acknowledgments
We are grateful the Center for AIDS Research Biostatistical Core for statistical consultation and the Partners HealthCare Research Patient Data Registry group for facilitating use of their database.
Disclaimer. The funding sources had no role in the design and conduct
of the study.
Financial support. This work was supported by the National Institutes
of Health (NIAID AI 007433 [to S. S.], NIH K01 AI07310 [to V. A. T.], K24
DK080140 [to J. B. M.], and K24 DK064545 [to S. K. G.]).
Potential conicts of interest. S. K. G. received research support from
Amgen, Bristol-Myers Squibb, Gilead, and Theratechnologies. He also
served as a consultant to Aileron.
All authors have submitted the ICMJE Form for Disclosure of Potential
Conicts of Interest. Conicts that the editors consider relevant to the content of the manuscript have been disclosed.

References
1. van Sighem AI, Gras LA, Reiss P, et al. Life expectancy of recently diagnosed asymptomatic HIV-infected patients approaches that of uninfected individuals. AIDS 2010; 24:152735.
2. Lewden C, Bouteloup V, De Wit S, et al. All-cause mortality in treated
HIV-infected adults with CD4 500/mm3 compared with the general
population: evidence from a large European observational cohort collaboration. Int J Epidemiol 2012; 41:43345.
3. Sackoff JE, Hanna DB, Pfeiffer MR, et al. Causes of death among persons with AIDS in the era of highly active antiretroviral therapy:
New York City. Ann Intern Med 2006; 145:397406.
4. Neuhaus J, Angus B, Kowalska JD, et al. Risk of all-cause mortality associated with nonfatal AIDS and serious non-AIDS events among
adults infected with HIV. AIDS 2010; 24:697706.

Aspirin Use and HIV

OFID

Downloaded from http://ofid.oxfordjournals.org/ by guest on October 20, 2014

CONCLUSIONS

5. Currier JS, Taylor A, Boyd F, et al. Coronary heart disease in HIVinfected individuals. J Acquir Immune Dec Syndr 2003; 33:50612.
6. Triant VA, Lee H, Hadigan C, et al. Increased acute myocardial infarction rates and cardiovascular risk factors among patients with human
immunodeciency virus disease. J Clin Endocrinol Metab 2007; 92:
250612.
7. Lang S, Mary-Krause M, Cotte L, et al. Increased risk of myocardial
infarction in HIV-infected patients in France, relative to the general
population. AIDS 2010; 24:122830.
8. Klein D, Hurley LB, Quesenberry CP, et al. Do protease inhibitors
increase the risk for coronary heart disease in patients with HIV-1 infection? J Acquir Immune Dec Syndr 2002; 30:4717.
9. Freiberg MS, Chang CC, Kuller LH, et al. HIV infection and the risk of
acute myocardial infarction. JAMA Intern Med 2013; 173:61422.
10. Friis-Moller N, Weber R, Reiss P, et al. Cardiovascular disease risk factors in HIV patientsassociation with antiretroviral therapy. Results
from the DAD study. AIDS 2003; 17:117993.
11. Holmberg SD, Moorman AC, Williamson JM, et al. Protease inhibitors
and cardiovascular outcomes in patients with HIV-1. Lancet 2002; 360:
17478.
12. El-Sadr WM, Lundgren JD, Neaton JD, et al. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med 2006; 355:228396.
13. Triant VA, Regan S, Lee H, et al. Association of immunologic and virologic factors with myocardial infarction rates in a US healthcare system.
J Acquir Immune Dec Syndr 2010; 55:6159.
14. Steering Committee of the Physicians Health Study Research Group.
Final report on the aspirin component of the ongoing Physicians
Health Study. N Engl J Med 1989; 321:12935.
15. Baigent C, Blackwell L, Collins R, et al. Aspirin in the primary and secondary prevention of vascular disease: collaborative meta-analysis of individual participant data from randomised trials. Lancet 2009;
373:184960.
16. Seshasai SR, Wijesuriya S, Sivakumaran R, et al. Effect of aspirin on vascular and nonvascular outcomes: meta-analysis of randomized controlled trials. Arch Intern Med 2012; 172:20916.
17. Antithrombotic Trialists Collaboration. Collaborative meta-analysis of
randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ 2002;
324:7186.
18. Triant VA, Brown TT, Lee H, et al. Fracture prevalence among human
immunodeciency virus (HIV)-infected versus non-HIV-infected patients in a large U.S. healthcare system. J Clin Endocrinol Metab
2008; 93:3499504.
19. Triant VA, Meigs JB, Grinspoon SK. Association of C-reactive protein
and HIV infection with acute myocardial infarction. J Acquir Immune
Dec Syndr 2009; 51:26873.
20. Zeng QT, Goryachev S, Weiss S, et al. Extracting principal diagnosis, comorbidity and smoking status for asthma research: evaluation of a natural language processing system. BMC Med Inform Decis Mak 2006;
6:30.
21. Regan S, Meigs JB, Grinspoon SK, et al. Determinants of smoking and
quitting in an HIV cohort using a validated natural language processing
tool. In: Abstract P168 of the Epidemiology and Prevention and Nutrition,
Physical Activity and Metabolism meeting. San Diego, 1316 March 2012.
22. Burkholder GA, Tamhane AR, Salinas JL, et al. Underutilization of
aspirin for primary prevention of cardiovascular disease among HIVinfected patients. Clin Infect Dis 2012; 55:15507.
23. Tornero C, Ventura A, Mafe M. Aspirin is indicated for primary prevention of cardiovascular events in HIV-infected patients. J Acquir Immune Dec Syndr 2010; 54:560.
24. Aberg JA, Gallant JE, Ghanem KG, et al. Primary care guidelines for the
management of persons infected with HIV: 2013 Update by the HIV
Medicine Association of the Infectious Diseases Society of America.
Clin Infect Dis 2014; 58:110.
25. Stone NJ, Robinson J, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic
cardiovascular risk in adults: a report of the American College of

26.
27.

28.

29.

Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2014; 129(25 Suppl 2):S145.
Momplaisir F, Mounzer K, Long JA. Preventive cancer screening practices in HIV-positive patients. AIDS Care 2014; 26:8794.
Lichtenstein KA, Armon C, Buchacz K, et al. Provider compliance with
guidelines for management of cardiovascular risk in HIV-infected patients. Prev Chronic Dis 2013; 10:E10.
Stafford RS, Monti V, Ma J. Underutilization of aspirin persists in US
ambulatory care for the secondary and primary prevention of cardiovascular disease. PLoS Med 2005; 2:e353.
Bartolucci AA, Tendera M, Howard G. Meta-analysis of multiple primary prevention trials of cardiovascular events using aspirin. Am J Cardiol 2011; 107:1796801.

30. Davi G, Patrono C. Platelet activation and atherothrombosis. N Engl J

Med 2007; 357:248294.
31. Holme PA, Muller F, Solum NO, et al. Enhanced activation of platelets
with abnormal release of RANTES in human immunodeciency virus
type 1 infection. FASEB J 1998; 12:7989.
32. Satchell CS, Cotter AG, OConnor EF, et al. Platelet function and HIV: a
case-control study. AIDS 2010; 24:64957.
33. OBrien M, Montenont E, Hu L, et al. Aspirin attenuates platelet activation and immune activation in HIV-1-infected subjects on antiretroviral
therapy: a pilot study. J Acquir Immune Dec Syndr 2008; 63:2808.
34. Chow FC, Regan S, Feske S, et al. Comparison of ischemic stroke incidence in HIV-infected and non-HIV-infected patients in a US health
care system. J Acquir Immune Dec Syndr 2012; 60:3518.

Downloaded from http://ofid.oxfordjournals.org/ by guest on October 20, 2014

OFID

Suchindran et al.